HHS Public Access Author manuscript Author Manuscript
Pediatr Res. Author manuscript; available in PMC 2017 August 11. Published in final edited form as: Pediatr Res. 2017 August ; 82(2): 201–208. doi:10.1038/pr.2017.34.
Peripubertal dietary flavonol and lignan intake and age at menarche in a longitudinal cohort of girls Nancy A. Mervish1,*, Susan L. Teitelbaum1, Ashley Pajak1, Gayle C. Windham2, Susan M. Pinney3, Lawrence H. Kushi4, Frank M. Biro5, and Mary S. Wolff1 on behalf of the Breast Cancer and Environment Research Programs 1Department
Author Manuscript
of Environmental Medicine and Public Health, Icahn School of Medicine at Mount Sinai School, New York, NY
2California
Department of Public Health, Division of Environmental and Occupational Disease Control, Richmond, CA 3Department
of Environmental Health, University of Cincinnati College of Medicine, Cincinnati,
OH 4Division
of Research, Kaiser Permanente Northern California, Oakland, CA
5Department
of Pediatrics, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH
Abstract Author Manuscript
Background—Dietary phytoestrogens may alter hormonal activity in childhood. Flavonols and lignans are the most prevalent phytoestrogens in the Western diet. We examined whether higher intake of flavonols and lignans were associated with later age at menarche in a prospective study of young girls. Methods—1044 girls ages 6–8 years (mean 7.3 yr) with 2–4 24 hour dietary recalls during their baseline year were followed for 11 years until attainment of menarche in the Breast Cancer and Environment Research Project (BCERP). Associations of age at menarche with quintiles of phytoestrogens were assessed using hazard ratios (HR) and 95% confidence intervals (CI) from Cox proportional hazards models, controlling for body mass index and other covariates.
Author Manuscript
Results—The highest quintile of flavonol intake was associated with a later age at menarche, compared to the lowest quintile (adjusted Hazard Ratio (HR): 0.80, 95% CI: (0.66–1.00). For lignans, there was a later age in overweight girls (HR: 0.56 95% CI=0.40–0.80). Conclusion—These dietary bioactives may reflect a healthy diet and foods high in phytoestrogens may influence the timing of menarche.
*
Corresponding Author: Nancy Mervish PhD, Department of Environmental Medicine and Public Health, Icahn School of Medicine at Mount Sinai. 1 Gustave Levy Place, NY, NY 10032, [ [email protected]], 212-824-7004. Financial Disclosure: The authors have no financial relationships relevant to this article to disclose.
Mervish et al.
Page 2
Author Manuscript
Introduction
Author Manuscript
Early menarche has been associated with breast, endometrial and ovarian cancer, possibly because it may lead to increased lifetime estrogen exposure (1). Although timing of menarche is under genetic control (2), environmental factors may also play an important role. Childhood nutrition affects adolescent growth and pubertal development (3). In addition to overall nutrition status, specific dietary components during childhood may influence timing of puberty and menarche. Fiber (4, 5) has mainly been associated with a later age at menarche whereas animal fats (6) and protein (7), dairy (8, 9), are associated with earlier menarche, although not consistently across studies (10–13). Longitudinal studies of phytoestrogens, including soy formula, dietary and urinary isoflavones, enterolactone and flavonols have shown later pubertal development (14–17). Reasons for differences may include prospective versus cross-sectional study design and timing of diet collection. Understanding the relationship between diet and menarche can help identify modifiable risk factors for disease prevention later in life.
Author Manuscript
Phytoestrogens are naturally occurring plant-derived polyphenols found in many foods that have been postulated to have beneficial health effects (18). The main groups of phytoestrogens are lignans, isoflavones and coumestans. Numerous epidemiologic studies have suggested that dietary intake of phytoestrogens, particularly isoflavones, play a role in preventing certain types of hormonally-dependent cancers, including breast and prostate (19, 20). Mechanisms of the protective effect are not fully understood, although it may be explained by the competition of phytoestrogens with endogenous estradiol for hormone receptors or an inhibitory effect on aromatase activity and antioxidant properties (21). The intake of soy products that are the major contributors to isoflavone intake is low in Western societies, whereas the intake of Flavonols and lignans is higher in the Western diet. Flavonols and lignans have weak estrogenic activity (22) and they have been associated with estrogen dependent outcomes, such as puberty, fecundity, menarche and ovarian hormone profiles (19, 23, 24). Identifying the major sources of foods eaten in young girls that contain high concentrations of these micronutrients may allow for targeted recommendations for health care providers. Our previous work has shown these phytoestrogens to be associated with a delay in girls’ breast development (14), so we hypothesize that they also may be associated with later menarche. We examined whether increased intake of flavonols and lignans before puberty are associated with later age at menarche in a prospective study of young girls.
Methods Author Manuscript
This project was part of the NIEHS/NCI Breast Cancer and the Environment Research Project (BCERP). BCERP is an observational longitudinal study of pubertal development of 1239 girls enrolled at 6–8 years old. Enrollment occurred between 2004–2007 at 3 sites: Icahn School of Medicine at Mount Sinai (NYC), which recruited through clinics, schools, and neighborhood centers in East Harlem, New York, with girls seen annually; Cincinnati Children’s Hospital/University of Cincinnati (Cincinnati), which recruited through schools in the Cincinnati metropolitan area and through the Breast Cancer Registry of Greater
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 3
Author Manuscript
Cincinnati, with girls seen semiannually; and Kaiser Permanente Northern California (KPNC), which recruited KPNC Health Plan members in the San Francisco Bay Area, with girls seen annually. The study was approved by the institutional review board at each site. Eligible girls had no underlying conditions affecting metabolism or growth. All sites obtained informed consent from parent or guardian and assent from girls. Parents or guardians of the participants identified girls’ race/ethnicity as Black, White, Asian, and ethnicity as Hispanic or non-Hispanic. At enrollment (baseline), and approximately annually for 11 years, parents or guardians were interviewed in person or answered a self-completed questionnaire, in either English or Spanish, about the children’s environmental exposures, physical activity, medical history, and demographics. Anthropometry and pubertal staging were assessed at each visit. Questions about the presence and date of first menarche were asked after the first follow-up year in NYC and KPNC and approximately 5 years after baseline in Cincinnati. A complete description of the study design is available in Biro et. al 2010 (25). We report on 1044 girls with 2–4 diet recalls provided within their first year of enrollment, those asked about menarche at least one time through 2014 (mean years of follow-up = 9).
Author Manuscript
Assessment of age at menarche Age at menarche was assigned based on questionnaire information provided primarily by the parent or guardian (N=795, 76%) or girl’s self-reported information (N=112, 11%). Girls who did not reach menarche were censored for statistical purposes at the age at their last visit or response (n=137,13%). Dietary Data
Author Manuscript
Dietary recall interviews were performed by trained dieticians at Cincinnati Center for Nutritional Research and Analysis using the NDSR,Minn, Minnesota. Multiple (2–4) 24hour recalls by telephone were conducted with caregivers during the first year to provide an average intake over a year’s time, as previously described (14). Average intakes over one year account for daily and seasonal variation in food consumed. Three recalls have been shown to reliably rank individuals on energy and most nutrients (26). The majority of girls had a complete set of 4 recalls (N=984); the remainder had either 2 or 3 recalls (N= 148 and 46, respectively). Recalls with either extremely low or high total energy consumption values (4000 k/cal per day) were excluded from analysis (N=18).
Author Manuscript
We developed a dietary database that captured flavonol and lignan contents of several hundred foods that were unavailable in standard dietary databases including NDSR (14). We assigned flavonol and lignan content (mg/100g) for 2319 individual foods and components reported, using published sources, including the USDA. Flavonol and lignan intake was calculated for each reported food by multiplying quantity consumed (g/day) and phytoestrogen content (mg/g). Total daily flavonol and lignan intake (mg/day) was averaged over the 2–4 calls for each girl. Statistical Analyses Baseline characteristics of the study population were examined by quintiles of flavonol and lignan intake. Due to skewed distributions, continuous variables are summarized as medians
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 4
Author Manuscript
and interquartile ranges (IQR). We tested for differences across quintiles of phytoestrogens by demographic characteristics using either chi-square test (categorical variables) or Kruskal-Wallis (continuous variables). We computed the amount of total flavonol and lignans in a standard serving size of foods selected as being high in these phytoestrogens, and then compiled the frequency of girls in our study who ate these foods (Tables 2 and 3).
Author Manuscript
Longitudinal analyses, examining associations between baseline flavonol and lignan intake and menarche, utilized reports of menarche over 11 years of annual follow up visits. We used Cox proportional-hazards models to calculate the hazard ratios (HR) and 95% confidence intervals (CIs) for associations between phytoestrogen levels and menarche. Proportional hazards assumptions were met in all of our analyses. Censoring occurred at the age of first menstrual period or age at last visit if menarche had not been reported. We examined flavonol and lignan intake using quintiles both as absolute values and calorie adjusted using the nutrient density method to remove variation caused by total energy intake (26). Associations were similar; calorie adjusted are reported only. All analyses were performed using the Statistics Analysis System (SAS) statistical software, Version 9.3 (SAS Institute, Cary, NC).
Author Manuscript
Confounding by potential covariates was assessed for the following variables: self-declared race/ethnicity, parent or guardian education, total daily average calories, fiber intake and BMI percentile (BMI%) dichotomized at the 85th percentile which defines overweight (specific for age and gender) according to the CDC(27).We examined BMI % separately at two timepoints: at baseline and at visit closest to menarche. These variables were chosen because they have been associated with both menarche and diet in the literature and based on a conceptual model of diet and pubertal development. We included covariates that altered the estimate by more than 10%. We did not include site because of its collinearity with race and BMI%. Final models were adjusted for self-declared race/ethnicity and baseline BMI %. Tests for trend across phytoestrogen quintiles were performed by including flavonol and lignan quintile ordinals as continuous variables. P-Contrasts were obtained from the contrast statement comparing the hazard ratio of the first quintile to the fifth quintile. We obtained adjusted median pubertal ages for flavonol and lignan intake using survival function estimates from the baseline survivor function of multivariable adjusted Cox models with quintiles as a strata statement. We evaluated race and baseline BMI% as effect modifiers by adding interaction terms to the model (flavonol or lignan quintile × BMI% dichotomized at the 85th percentile and flavonol or lignan quintile × race). If there was a significant interaction either race or BMI% we examined stratum specific models.
Author Manuscript
Results Table 1 presents the demographic characteristics of the study population at baseline.All girls consumed at least one food with flavonol and lignan content. The median intake of flavonols was 6.0 mg/day, median of lignans was 0.40 mg/day. Flavonols and lignans are found in lower concentrations in more frequently consumed foods such as potatoes, grains, cereals, tomato products, coffee, and (Tables 2 and 3), compared to foods like kale and arugula. The average intake consumed in our population for both flavonol and lignans was a full serving per day for some of the foods (e.g., grapefruit juice, pears, strawberries).
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 5
Author Manuscript
Tables 4 and 5 show the baseline characteristics of 1044 participants according to quintiles of flavonol and lignan intake. Girls with higher flavonol and lignan intake were more likely to be white, have more educated parents/guardians and consume more calories. At the visit when they were last seen 907 (87%) of the girls had reported a first menstrual period. The unadjusted median age at menarche in our sample was 12.2 years (IQR: 11.4–13.1 years). This was the same as the median of the total cohort of 1089 reporting age at menarche (median age 12.3 years (IQR: 11.4–13.2 years), which included girls without diet data.
Author Manuscript
The highest quintile of flavonol intake was associated with a later age at menarche, compared to the lowest quintile (adjusted HR: 0.80, 95% CI: (0.66–1.00),(Table 6). Compared to the lowest lignan intake (median 0.14 mg/day), highest intake (median 0.88 mg/day) was also suggestive of an association with later menarche (adjusted HR: 0.84, 95% CI: (0.66–1.04) (Table 6). We additionally adjusted for average fiber intake and estimates were unchanged (data not shown). The BMI-lignan interaction was significant in the test for effect modification (p for interaction=0.02). As shown in table 7, the association with later menarche was restricted to girls with higher BMI (HR: 0.56 95% CI=0.40–0.80). There was no evidence of interaction by BMI with flavonols nor was there interaction of the association of lignans or flavonols and menarche by race (data not shown).
Discussion
Author Manuscript
Our findings suggest that specific childhood dietary factors may influence the pubertal trajectory. In this longitudinal cohort, flavonol consumption was associated with later age at menarche, in keeping with our hypothesis. Similar relationships were observed in this cohort between increased intake of these phytoestrogens and later age at breast development, which typically precedes menarche by 2–3 years. Association of lignans was restricted to overweight and obese girls, while flavonol associations were not modified by BMI. There was a six month delay in menarche among overweight girls with high lignan consumption compared to those with low lignan intake. To put this age difference in perspective, 6 months is the difference seen in black and white girls in a number of studies (28). This present study is the only one that has examined the association between these two phytoestrogens, which are widespread in the Western diet, and age at menarche. Previous longitudinal and cross sectional studies have looked at soy and other dietary components and age at menarche, with mixed results, possibly due to when during childhood diet was assessed and type of database used (10, 11, 13, 16)
Author Manuscript
Our findings are particularly noteworthy, given that flavonol and lignan intake in this cohort is relatively low compared to a few studies reporting these phytoestrogens in US adults (29, 30). Intake of phytoestrogens is usually lower in children, especially in an urban low SES population (14). Flaxseeds are rich sources of lignans (80mg per 1oz serving size), although they were not eaten in our cohort. In contrast, flavonols and lignans are found in much lower concentrations in many more frequently consumed foods such as potatoes, grains, cereals,, strawberries, coffee, tea and wine. Girls in this cohort consumed a serving size per day of some of the foods with the highest content of flavonol and lignan, (tables 2 and 3). Isoflavones, important phytoestrogens associated with reduced risk of breast cancer, and a later onset of puberty, come from different types of foods, compared to flavonols and
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 6
Author Manuscript
lignans. Isoflavones are found in soy foods (soybeans, tofu, soy milk, soy nuts), which are infrequently eaten in the US compared to the foods high in flavonols and lignans. We did not examine these in our study. While some foods high in lignans, flaxseeds and spinach also contain isoflavones, the concentration of isoflavones in these foods is extremely low when compared to soybeans.
Author Manuscript
BMI has been shown to be a strong predictor of early menarche and early pubertal development (31). We saw that overweight/obesity modified the relationship between lignan intake and age at menarche. This finding is consistent with evidence of possible mechanisms that lignans regulate adipogenesis through the modulation of the estrogen receptor signaling pathway in both animal and human studies (21) and that phytoestrogens activate the PPARγ receptor which controls many genes associated with fat and energy metabolism (32). Lignans are metabolized to enterolactone and enterodiol by gut microbiota in the colon, and a few human studies have shown that higher consumption of lignans was associated with lower overall fat mass (33). Enterolactone is also an aromatase inhibitor, which could reduce estrogen levels essential for pubertal changes.
Author Manuscript
Studies of menarche have measured diet at different time points of childhood, ranging from 1–3 years (34) to mid childhood (9 yrs) (8) and later (9–14yrs) (10). Evidence is mixed as to when the critical period may be for diet to impact the occurrence of menarche. Studies have shown associations between early childhood diet and growth and pubertal development (35, 36). Birth may be a critical window for environmental exposures related to pubertal development, as it is a time when the reproductive tract is susceptible to changes in the developmental process. Soy intake during infancy altered the timing of menarche (17). Rapid weight gain in infancy and early childhood is a determinant of timing of pubertal development and this reflects the importance of early “windows” during which nutrition can have long term consequences for growth and development (37). However, later “windows of susceptibility” may have a different influence. For example, a protective effect of dietary phytoestrogen intake on breast cancer risk was observed for relatively high isoflavone consumption before puberty (18) and during adolescence (38). Phytoestrogen exposure may need to occur in a critical window long before the outcome to have an influence.
Author Manuscript
Strengths of this study are its prospective nature, comprehensive diet assessment and reliable outcome. We measured diet when girls were 6–8 years of age, several years before menarche, strengthening conclusions of the observed relationship. Diet was assessed using an average of multiple 24 hour dietary recalls and utilized an extensive database created to assess flavonol and lignan intake. The idea that dietary intake may change over time has been studied and in children 9–18 years, dietary intake patterns that were assessed 5 years apart were stable over time (39). A main limitation is the method of assessing flavonol and lignan intake. Our database was developed using the currently known information of foods with flavonol and lignan content, which may not have had flavonol and lignan content of every food item reported in the Nutrition Data System for Research (NDSR, Minn, Minnesota) questionnaire. Another method of lignan exposure assessment is using urine biomarkers, however biomarkers have temporality concerns. In our previous report of this cohort for diet and breast development we showed low-moderate correlations between urine
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 7
Author Manuscript
enterolactone and lignan intake assessed through dietary records, and urinary enterolactone was associated with pubertal development (14, 25). We report an association between higher childhood consumption of flavonols and lignans and later age at menarche. These agents have known anti-hormonal and anti-obesogenic properties. It is possible that flavonols and lignans reflect a healthy diet, as they are derived mainly from fruits, vegetables and fiber in our cohort. However the associations remain after adjusting for fiber, which may be another surrogate for a healthy diet. While it is difficult to disentangle the individual effects of these phytoestrogens from fruits, vegetables and fiber, nonetheless they all may play a role in the observed pubertal delay. Our findings suggest that specific dietary bioactives in certain fruits and vegetables may have additional long term benefits and dietary modifications are of continued importance to young girls.
Author Manuscript
Acknowledgments We gratefully acknowledge our collaborators at the 3 medical centers involved in this research including Jessica Guiterrez, Rochelle Osborne, Lisa Boguski, Joel Forman, and Barbara Brenner (MSSM); Gayle Greenberg, Bob Bornschein (Cincinnati); Robert Hiatt, Louise Greenspan, Julie Deardorff (Kaiser Permanent Funding Source: This publication was made possible by Avon Foundation (New York, NY), the Breast Cancer and the Environment Research Program (BCERP) award numbers U01ES012770, U01ES012771, U01ES012800, U01ES012801, U01ES019435, U01ES019453, U01ES019454 and U01ES019457 from the National Institute of Environmental Health Sciences (NIEHS), Research Triangle Park, North Carolina, and the National Cancer Institute (NCI), Rockville, MD, P01ES009584, P30ES023514, P30ES023515 and P30ES0060696 from NIEHS, and UL1RR024131, UL1RR026314 and UL1RR029887 from the National Cancer for Research Resources (NCRR), Bethesda, MD. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIEHS or NCI, the National Institutes of Health, the Centers for Disease Control and Prevention, Atlanta, Georgia, or the California Department of Public Health.
Author Manuscript
Reference List
Author Manuscript
1. Rosner B, Colditz GA, Willett WC. Reproductive risk factors in a prospective study of breast cancer: the Nurses’ Health Study. Am J Epidemiol. 1994; 139:819–835. [PubMed: 8178795] 2. Wehkalampi K, Silventoinen K, Kaprio J, et al. Genetic and environmental influences on pubertal timing assessed by height. Am J Hum Biol. 2008; 20:417–423. [PubMed: 18293372] 3. Rogol AD, Clark PA, Roemmich JN. Growth and pubertal development in children and adolescents: effects of diet and physical activity. Am J Clin Nutr. 2000; 72:521S–528S. [PubMed: 10919954] 4. Hughes RE, Jones E. Intake of dietary fibre and the age of menarche. Ann Hum Biol. 1985; 12:325– 332. [PubMed: 2994546] 5. Koo MM, Rohan TE, Jain M, McLaughlin JR, Corey PN. A cohort study of dietary fibre intake and menarche. Public Health Nutr. 2002; 5:353–360. [PubMed: 12020388] 6. Merzenich H, Boeing H, Wahrendorf J. Dietary fat and sports activity as determinants for age at menarche. Am J Epidemiol. 1993; 138:217–224. [PubMed: 8356963] 7. Gunther AL, Karaolis-Danckert N, Kroke A, Remer T, Buyken AE. Dietary protein intake throughout childhood is associated with the timing of puberty. J Nutr. 2010; 140:565–571. [PubMed: 20042466] 8. Wiley AS. Milk intake and total dairy consumption: associations with early menarche in NHANES 1999–2004. PLoS One. 2011; 6:e14685. [PubMed: 21347271] 9. Ramezani TF, Moslehi N, Asghari G, et al. Intake of dairy products, calcium, magnesium, and phosphorus in childhood and age at menarche in the Tehran Glucose and Lipid Study. PLoS One. 2013; 8:e57696. [PubMed: 23451261] 10. Carwile JL, Willett WC, Wang M, et al. Milk Consumption after Age 9 Years Does Not Predict Age at Menarche. J Nutr. 2015; 145:1900–8. [PubMed: 26136590] 11. Moisan J, Meyer F, Gingras S. Diet and age at menarche. Cancer Causes Control. 19901:149–154. Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 8
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
12. Cheng G, Buyken AE, Shi L, et al. Nutr Rev. 2012; 70:133–152. [PubMed: 22364156] 13. Segovia-Siapco G, Pribis P, Messina M, Oda K, Sabate J. Is soy intake related to age at onset of menarche? A cross-sectional study Nutr J. 2014; 13:54.doi: 10.1186/1475-2891-13-54.:54-13 [PubMed: 24889551] 14. Mervish NA, Gardiner EW, Galvez MP, et al. Dietary flavonol intake is associated with age of puberty in a longitudinal cohort of girls. Nutr Res. 2013; 33:534–542. [PubMed: 23827127] 15. Wolff MS, Britton JA, Boguski L, et al. Environmental exposures and puberty in inner-city girls. Environ Res. 2008; 107:393–400. [PubMed: 18479682] 16. Cheng G, Remer T, Prinz-Langenohl R, et al. Relation of isoflavones and fiber intake in childhood to the timing of puberty. Am J Clin Nutr. 2010; 92:556–564. [PubMed: 20631200] 17. Adgent MA, Daniels JL, Rogan WJ, et al. Early-life soy exposure and age at menarche. Paediatr Perinat Epidemiol. 2012; 26:163–175. [PubMed: 22324503] 18. Adlercreutz H. Phytoestrogens and breast cancer. J Steroid Biochem Mol Biol. 2002; 83:113–118. [PubMed: 12650707] 19. Buck K, Zaineddin AK, Vrieling A, Linseisen J, Chang-Claude J. Meta-analyses of lignans and enterolignans in relation to breast cancer risk. Am J Clin Nutr. 2010; 92:141–153. [PubMed: 20463043] 20. Giles GG. Epidemiological investigation of prostate cancer. Methods Mol Med. 2003; 81:1–19. [PubMed: 12725111] 21. Adlercreutz H, Bannwart C, Wahala K, et al. Inhibition of human aromatase by mammalian lignans and isoflavonoid phytoestrogens in childhood and age at menarche in the Tehran Lipid and Glucose Study. J Steroid Biochem. 1993; 44:147–153. Mol Biol. 22. Setchell KD. Phytoestrogens: the biochemistry, physiology, and implications for human health of soy isoflavones. Am J Clin Nutr. 1998; 68:1333S–1346S. [PubMed: 9848496] 23. Linseisen J, Piller R, Hermann S, Chang-Claude J. Dietary phytoestrogen intake and premenopausal breast cancer risk in a German case-control study. Int J Cancer. 2004; 110:284– 290. [PubMed: 15069695] 24. Mumford SL, Sundaram R, Schisterman EF, et al. Higher urinary lignan concentrations in women but not men are positively associated with shorter time to pregnancy. J Nutr. 2014; 144:352–358. [PubMed: 24401816] 25. Biro FM, Galvez MP, Greenspan LC, et al. Pubertal Assessment Method and Baseline Characteristics in a Mixed Longitudinal Study of Girls. Pediatrics. 2010; 126(3):e583–90. [PubMed: 20696727] 26. Willett W, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol. 1986; 124:17–27. [PubMed: 3521261] 27. CDC. Overweight and Obesity. 2015. www.cdc.gov/obesity/childhood/defining.html 28. Euling SY, Selevan SG, Pescovitz OH, Skakkebaek NE. Role of environmental factors in the timing of puberty. Pediatrics. 2008; 121(Suppl 3):S167–71. S167-S171. DOI: 10.1542/peds. 2007-1813C [PubMed: 18245510] 29. Sampson L, Rimm E, Hollman PC, de Vries JH, Katan MB. Flavonol and flavone intakes in US health professionals. J Am Diet Assoc. 2002; 102:1414–1420. [PubMed: 12396158] 30. Fink BN, Steck SE, Wolff MS, Kabat GC, Gammon MD. Construction of a flavonoid database for assessing intake in a population-based sample of women on Long Island, New York. Nutr Cancer. 2006; 56:57–66. [PubMed: 17176218] 31. Biro FM, Khoury P, Morrison JA. Influence of obesity on timing of puberty. Int J Androl. 2006; 29:272–277. [PubMed: 16371114] 32. Jungbauer A, Medjakovic S. Phytoestrogens and the metabolic syndrome. J Steroid Biochem Mol Biol. 2014; 139:277–89. Epub@2013 Jan 11.:277–289. DOI: 10.1016/j.jsbmb.2012.12.009 [PubMed: 23318879] 33. Hu Y, Song Y, Franke AA, Hu FB, van Dam RM. A Prospective Investigation of the Association Between Urinary Excretion of Dietary Lignan Metabolites and Weight Change in US Women. Am J Epidemiol. 2015; 182:503–511. [PubMed: 26290574]
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 9
Author Manuscript Author Manuscript
34. Berkey CS, Gardner JD, Frazier AL, Colditz GA. Relation of childhood diet and body size to menarche and adolescent growth in girls. Am J Epidemiol. 2000; 152:446–452. [PubMed: 10981459] 35. dos Santos Silva I, De Stavola BL, Mann V, Kuh D, Hardy R, Wadsworth ME. Prenatal factors, childhood growth trajectories and age at menarche. Int J Epidemiol. 2002; 31:405–412. [PubMed: 11980804] 36. Sloboda DM, Hart R, Doherty DA, Pennell CE, Hickey M. Age at menarche: Influences of prenatal and postnatal growth. J Clin Endocrinol Metab. 2007; 92:46–50. [PubMed: 17062767] 37. Ong KK, Ahmed ML, Emmett PM, Preece MA, Dunger DB. Association between postnatal catchup growth and obesity in childhood: prospective cohort study. BMJ. 2000; 320:967–971. [PubMed: 10753147] 38. Thanos J, Cotterchio M, Boucher BA, Kreiger N, Thompson LU. Adolescent dietary phytoestrogen intake and breast cancer risk (Canada). Cancer Causes Control. 2006; 17:1253–1261. [PubMed: 17111256] 39. Cutler GJ, Flood A, Hannan P, Neumark-Sztainer D. Major patterns of dietary intake in adolescents and their stability. J Nutr. 2009; 139:323–328. [PubMed: 19091799]
Author Manuscript Author Manuscript Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 10
Table 1
Author Manuscript
Baseline Characteristics of 1044 BCERP participants, 2004–2014. Characteristic Mean (Std) Baseline age (years)
7.3 (.68)
Child race/ethnicity
N
%
Black
316
30
Hispanic
316
30
White
357
34
Asian
55
5
≤ High school
296
29
> High school
725
71
High school
35 29 36
High school
30 35 35
Pediatr Res. Author manuscript; available in PMC 2017 August 11. Published in final edited form as: Pediatr Res. 2017 August ; 82(2): 201–208. doi:10.1038/pr.2017.34.
Peripubertal dietary flavonol and lignan intake and age at menarche in a longitudinal cohort of girls Nancy A. Mervish1,*, Susan L. Teitelbaum1, Ashley Pajak1, Gayle C. Windham2, Susan M. Pinney3, Lawrence H. Kushi4, Frank M. Biro5, and Mary S. Wolff1 on behalf of the Breast Cancer and Environment Research Programs 1Department
Author Manuscript
of Environmental Medicine and Public Health, Icahn School of Medicine at Mount Sinai School, New York, NY
2California
Department of Public Health, Division of Environmental and Occupational Disease Control, Richmond, CA 3Department
of Environmental Health, University of Cincinnati College of Medicine, Cincinnati,
OH 4Division
of Research, Kaiser Permanente Northern California, Oakland, CA
5Department
of Pediatrics, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH
Abstract Author Manuscript
Background—Dietary phytoestrogens may alter hormonal activity in childhood. Flavonols and lignans are the most prevalent phytoestrogens in the Western diet. We examined whether higher intake of flavonols and lignans were associated with later age at menarche in a prospective study of young girls. Methods—1044 girls ages 6–8 years (mean 7.3 yr) with 2–4 24 hour dietary recalls during their baseline year were followed for 11 years until attainment of menarche in the Breast Cancer and Environment Research Project (BCERP). Associations of age at menarche with quintiles of phytoestrogens were assessed using hazard ratios (HR) and 95% confidence intervals (CI) from Cox proportional hazards models, controlling for body mass index and other covariates.
Author Manuscript
Results—The highest quintile of flavonol intake was associated with a later age at menarche, compared to the lowest quintile (adjusted Hazard Ratio (HR): 0.80, 95% CI: (0.66–1.00). For lignans, there was a later age in overweight girls (HR: 0.56 95% CI=0.40–0.80). Conclusion—These dietary bioactives may reflect a healthy diet and foods high in phytoestrogens may influence the timing of menarche.
*
Corresponding Author: Nancy Mervish PhD, Department of Environmental Medicine and Public Health, Icahn School of Medicine at Mount Sinai. 1 Gustave Levy Place, NY, NY 10032, [ [email protected]], 212-824-7004. Financial Disclosure: The authors have no financial relationships relevant to this article to disclose.
Mervish et al.
Page 2
Author Manuscript
Introduction
Author Manuscript
Early menarche has been associated with breast, endometrial and ovarian cancer, possibly because it may lead to increased lifetime estrogen exposure (1). Although timing of menarche is under genetic control (2), environmental factors may also play an important role. Childhood nutrition affects adolescent growth and pubertal development (3). In addition to overall nutrition status, specific dietary components during childhood may influence timing of puberty and menarche. Fiber (4, 5) has mainly been associated with a later age at menarche whereas animal fats (6) and protein (7), dairy (8, 9), are associated with earlier menarche, although not consistently across studies (10–13). Longitudinal studies of phytoestrogens, including soy formula, dietary and urinary isoflavones, enterolactone and flavonols have shown later pubertal development (14–17). Reasons for differences may include prospective versus cross-sectional study design and timing of diet collection. Understanding the relationship between diet and menarche can help identify modifiable risk factors for disease prevention later in life.
Author Manuscript
Phytoestrogens are naturally occurring plant-derived polyphenols found in many foods that have been postulated to have beneficial health effects (18). The main groups of phytoestrogens are lignans, isoflavones and coumestans. Numerous epidemiologic studies have suggested that dietary intake of phytoestrogens, particularly isoflavones, play a role in preventing certain types of hormonally-dependent cancers, including breast and prostate (19, 20). Mechanisms of the protective effect are not fully understood, although it may be explained by the competition of phytoestrogens with endogenous estradiol for hormone receptors or an inhibitory effect on aromatase activity and antioxidant properties (21). The intake of soy products that are the major contributors to isoflavone intake is low in Western societies, whereas the intake of Flavonols and lignans is higher in the Western diet. Flavonols and lignans have weak estrogenic activity (22) and they have been associated with estrogen dependent outcomes, such as puberty, fecundity, menarche and ovarian hormone profiles (19, 23, 24). Identifying the major sources of foods eaten in young girls that contain high concentrations of these micronutrients may allow for targeted recommendations for health care providers. Our previous work has shown these phytoestrogens to be associated with a delay in girls’ breast development (14), so we hypothesize that they also may be associated with later menarche. We examined whether increased intake of flavonols and lignans before puberty are associated with later age at menarche in a prospective study of young girls.
Methods Author Manuscript
This project was part of the NIEHS/NCI Breast Cancer and the Environment Research Project (BCERP). BCERP is an observational longitudinal study of pubertal development of 1239 girls enrolled at 6–8 years old. Enrollment occurred between 2004–2007 at 3 sites: Icahn School of Medicine at Mount Sinai (NYC), which recruited through clinics, schools, and neighborhood centers in East Harlem, New York, with girls seen annually; Cincinnati Children’s Hospital/University of Cincinnati (Cincinnati), which recruited through schools in the Cincinnati metropolitan area and through the Breast Cancer Registry of Greater
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 3
Author Manuscript
Cincinnati, with girls seen semiannually; and Kaiser Permanente Northern California (KPNC), which recruited KPNC Health Plan members in the San Francisco Bay Area, with girls seen annually. The study was approved by the institutional review board at each site. Eligible girls had no underlying conditions affecting metabolism or growth. All sites obtained informed consent from parent or guardian and assent from girls. Parents or guardians of the participants identified girls’ race/ethnicity as Black, White, Asian, and ethnicity as Hispanic or non-Hispanic. At enrollment (baseline), and approximately annually for 11 years, parents or guardians were interviewed in person or answered a self-completed questionnaire, in either English or Spanish, about the children’s environmental exposures, physical activity, medical history, and demographics. Anthropometry and pubertal staging were assessed at each visit. Questions about the presence and date of first menarche were asked after the first follow-up year in NYC and KPNC and approximately 5 years after baseline in Cincinnati. A complete description of the study design is available in Biro et. al 2010 (25). We report on 1044 girls with 2–4 diet recalls provided within their first year of enrollment, those asked about menarche at least one time through 2014 (mean years of follow-up = 9).
Author Manuscript
Assessment of age at menarche Age at menarche was assigned based on questionnaire information provided primarily by the parent or guardian (N=795, 76%) or girl’s self-reported information (N=112, 11%). Girls who did not reach menarche were censored for statistical purposes at the age at their last visit or response (n=137,13%). Dietary Data
Author Manuscript
Dietary recall interviews were performed by trained dieticians at Cincinnati Center for Nutritional Research and Analysis using the NDSR,Minn, Minnesota. Multiple (2–4) 24hour recalls by telephone were conducted with caregivers during the first year to provide an average intake over a year’s time, as previously described (14). Average intakes over one year account for daily and seasonal variation in food consumed. Three recalls have been shown to reliably rank individuals on energy and most nutrients (26). The majority of girls had a complete set of 4 recalls (N=984); the remainder had either 2 or 3 recalls (N= 148 and 46, respectively). Recalls with either extremely low or high total energy consumption values (4000 k/cal per day) were excluded from analysis (N=18).
Author Manuscript
We developed a dietary database that captured flavonol and lignan contents of several hundred foods that were unavailable in standard dietary databases including NDSR (14). We assigned flavonol and lignan content (mg/100g) for 2319 individual foods and components reported, using published sources, including the USDA. Flavonol and lignan intake was calculated for each reported food by multiplying quantity consumed (g/day) and phytoestrogen content (mg/g). Total daily flavonol and lignan intake (mg/day) was averaged over the 2–4 calls for each girl. Statistical Analyses Baseline characteristics of the study population were examined by quintiles of flavonol and lignan intake. Due to skewed distributions, continuous variables are summarized as medians
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 4
Author Manuscript
and interquartile ranges (IQR). We tested for differences across quintiles of phytoestrogens by demographic characteristics using either chi-square test (categorical variables) or Kruskal-Wallis (continuous variables). We computed the amount of total flavonol and lignans in a standard serving size of foods selected as being high in these phytoestrogens, and then compiled the frequency of girls in our study who ate these foods (Tables 2 and 3).
Author Manuscript
Longitudinal analyses, examining associations between baseline flavonol and lignan intake and menarche, utilized reports of menarche over 11 years of annual follow up visits. We used Cox proportional-hazards models to calculate the hazard ratios (HR) and 95% confidence intervals (CIs) for associations between phytoestrogen levels and menarche. Proportional hazards assumptions were met in all of our analyses. Censoring occurred at the age of first menstrual period or age at last visit if menarche had not been reported. We examined flavonol and lignan intake using quintiles both as absolute values and calorie adjusted using the nutrient density method to remove variation caused by total energy intake (26). Associations were similar; calorie adjusted are reported only. All analyses were performed using the Statistics Analysis System (SAS) statistical software, Version 9.3 (SAS Institute, Cary, NC).
Author Manuscript
Confounding by potential covariates was assessed for the following variables: self-declared race/ethnicity, parent or guardian education, total daily average calories, fiber intake and BMI percentile (BMI%) dichotomized at the 85th percentile which defines overweight (specific for age and gender) according to the CDC(27).We examined BMI % separately at two timepoints: at baseline and at visit closest to menarche. These variables were chosen because they have been associated with both menarche and diet in the literature and based on a conceptual model of diet and pubertal development. We included covariates that altered the estimate by more than 10%. We did not include site because of its collinearity with race and BMI%. Final models were adjusted for self-declared race/ethnicity and baseline BMI %. Tests for trend across phytoestrogen quintiles were performed by including flavonol and lignan quintile ordinals as continuous variables. P-Contrasts were obtained from the contrast statement comparing the hazard ratio of the first quintile to the fifth quintile. We obtained adjusted median pubertal ages for flavonol and lignan intake using survival function estimates from the baseline survivor function of multivariable adjusted Cox models with quintiles as a strata statement. We evaluated race and baseline BMI% as effect modifiers by adding interaction terms to the model (flavonol or lignan quintile × BMI% dichotomized at the 85th percentile and flavonol or lignan quintile × race). If there was a significant interaction either race or BMI% we examined stratum specific models.
Author Manuscript
Results Table 1 presents the demographic characteristics of the study population at baseline.All girls consumed at least one food with flavonol and lignan content. The median intake of flavonols was 6.0 mg/day, median of lignans was 0.40 mg/day. Flavonols and lignans are found in lower concentrations in more frequently consumed foods such as potatoes, grains, cereals, tomato products, coffee, and (Tables 2 and 3), compared to foods like kale and arugula. The average intake consumed in our population for both flavonol and lignans was a full serving per day for some of the foods (e.g., grapefruit juice, pears, strawberries).
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 5
Author Manuscript
Tables 4 and 5 show the baseline characteristics of 1044 participants according to quintiles of flavonol and lignan intake. Girls with higher flavonol and lignan intake were more likely to be white, have more educated parents/guardians and consume more calories. At the visit when they were last seen 907 (87%) of the girls had reported a first menstrual period. The unadjusted median age at menarche in our sample was 12.2 years (IQR: 11.4–13.1 years). This was the same as the median of the total cohort of 1089 reporting age at menarche (median age 12.3 years (IQR: 11.4–13.2 years), which included girls without diet data.
Author Manuscript
The highest quintile of flavonol intake was associated with a later age at menarche, compared to the lowest quintile (adjusted HR: 0.80, 95% CI: (0.66–1.00),(Table 6). Compared to the lowest lignan intake (median 0.14 mg/day), highest intake (median 0.88 mg/day) was also suggestive of an association with later menarche (adjusted HR: 0.84, 95% CI: (0.66–1.04) (Table 6). We additionally adjusted for average fiber intake and estimates were unchanged (data not shown). The BMI-lignan interaction was significant in the test for effect modification (p for interaction=0.02). As shown in table 7, the association with later menarche was restricted to girls with higher BMI (HR: 0.56 95% CI=0.40–0.80). There was no evidence of interaction by BMI with flavonols nor was there interaction of the association of lignans or flavonols and menarche by race (data not shown).
Discussion
Author Manuscript
Our findings suggest that specific childhood dietary factors may influence the pubertal trajectory. In this longitudinal cohort, flavonol consumption was associated with later age at menarche, in keeping with our hypothesis. Similar relationships were observed in this cohort between increased intake of these phytoestrogens and later age at breast development, which typically precedes menarche by 2–3 years. Association of lignans was restricted to overweight and obese girls, while flavonol associations were not modified by BMI. There was a six month delay in menarche among overweight girls with high lignan consumption compared to those with low lignan intake. To put this age difference in perspective, 6 months is the difference seen in black and white girls in a number of studies (28). This present study is the only one that has examined the association between these two phytoestrogens, which are widespread in the Western diet, and age at menarche. Previous longitudinal and cross sectional studies have looked at soy and other dietary components and age at menarche, with mixed results, possibly due to when during childhood diet was assessed and type of database used (10, 11, 13, 16)
Author Manuscript
Our findings are particularly noteworthy, given that flavonol and lignan intake in this cohort is relatively low compared to a few studies reporting these phytoestrogens in US adults (29, 30). Intake of phytoestrogens is usually lower in children, especially in an urban low SES population (14). Flaxseeds are rich sources of lignans (80mg per 1oz serving size), although they were not eaten in our cohort. In contrast, flavonols and lignans are found in much lower concentrations in many more frequently consumed foods such as potatoes, grains, cereals,, strawberries, coffee, tea and wine. Girls in this cohort consumed a serving size per day of some of the foods with the highest content of flavonol and lignan, (tables 2 and 3). Isoflavones, important phytoestrogens associated with reduced risk of breast cancer, and a later onset of puberty, come from different types of foods, compared to flavonols and
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 6
Author Manuscript
lignans. Isoflavones are found in soy foods (soybeans, tofu, soy milk, soy nuts), which are infrequently eaten in the US compared to the foods high in flavonols and lignans. We did not examine these in our study. While some foods high in lignans, flaxseeds and spinach also contain isoflavones, the concentration of isoflavones in these foods is extremely low when compared to soybeans.
Author Manuscript
BMI has been shown to be a strong predictor of early menarche and early pubertal development (31). We saw that overweight/obesity modified the relationship between lignan intake and age at menarche. This finding is consistent with evidence of possible mechanisms that lignans regulate adipogenesis through the modulation of the estrogen receptor signaling pathway in both animal and human studies (21) and that phytoestrogens activate the PPARγ receptor which controls many genes associated with fat and energy metabolism (32). Lignans are metabolized to enterolactone and enterodiol by gut microbiota in the colon, and a few human studies have shown that higher consumption of lignans was associated with lower overall fat mass (33). Enterolactone is also an aromatase inhibitor, which could reduce estrogen levels essential for pubertal changes.
Author Manuscript
Studies of menarche have measured diet at different time points of childhood, ranging from 1–3 years (34) to mid childhood (9 yrs) (8) and later (9–14yrs) (10). Evidence is mixed as to when the critical period may be for diet to impact the occurrence of menarche. Studies have shown associations between early childhood diet and growth and pubertal development (35, 36). Birth may be a critical window for environmental exposures related to pubertal development, as it is a time when the reproductive tract is susceptible to changes in the developmental process. Soy intake during infancy altered the timing of menarche (17). Rapid weight gain in infancy and early childhood is a determinant of timing of pubertal development and this reflects the importance of early “windows” during which nutrition can have long term consequences for growth and development (37). However, later “windows of susceptibility” may have a different influence. For example, a protective effect of dietary phytoestrogen intake on breast cancer risk was observed for relatively high isoflavone consumption before puberty (18) and during adolescence (38). Phytoestrogen exposure may need to occur in a critical window long before the outcome to have an influence.
Author Manuscript
Strengths of this study are its prospective nature, comprehensive diet assessment and reliable outcome. We measured diet when girls were 6–8 years of age, several years before menarche, strengthening conclusions of the observed relationship. Diet was assessed using an average of multiple 24 hour dietary recalls and utilized an extensive database created to assess flavonol and lignan intake. The idea that dietary intake may change over time has been studied and in children 9–18 years, dietary intake patterns that were assessed 5 years apart were stable over time (39). A main limitation is the method of assessing flavonol and lignan intake. Our database was developed using the currently known information of foods with flavonol and lignan content, which may not have had flavonol and lignan content of every food item reported in the Nutrition Data System for Research (NDSR, Minn, Minnesota) questionnaire. Another method of lignan exposure assessment is using urine biomarkers, however biomarkers have temporality concerns. In our previous report of this cohort for diet and breast development we showed low-moderate correlations between urine
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 7
Author Manuscript
enterolactone and lignan intake assessed through dietary records, and urinary enterolactone was associated with pubertal development (14, 25). We report an association between higher childhood consumption of flavonols and lignans and later age at menarche. These agents have known anti-hormonal and anti-obesogenic properties. It is possible that flavonols and lignans reflect a healthy diet, as they are derived mainly from fruits, vegetables and fiber in our cohort. However the associations remain after adjusting for fiber, which may be another surrogate for a healthy diet. While it is difficult to disentangle the individual effects of these phytoestrogens from fruits, vegetables and fiber, nonetheless they all may play a role in the observed pubertal delay. Our findings suggest that specific dietary bioactives in certain fruits and vegetables may have additional long term benefits and dietary modifications are of continued importance to young girls.
Author Manuscript
Acknowledgments We gratefully acknowledge our collaborators at the 3 medical centers involved in this research including Jessica Guiterrez, Rochelle Osborne, Lisa Boguski, Joel Forman, and Barbara Brenner (MSSM); Gayle Greenberg, Bob Bornschein (Cincinnati); Robert Hiatt, Louise Greenspan, Julie Deardorff (Kaiser Permanent Funding Source: This publication was made possible by Avon Foundation (New York, NY), the Breast Cancer and the Environment Research Program (BCERP) award numbers U01ES012770, U01ES012771, U01ES012800, U01ES012801, U01ES019435, U01ES019453, U01ES019454 and U01ES019457 from the National Institute of Environmental Health Sciences (NIEHS), Research Triangle Park, North Carolina, and the National Cancer Institute (NCI), Rockville, MD, P01ES009584, P30ES023514, P30ES023515 and P30ES0060696 from NIEHS, and UL1RR024131, UL1RR026314 and UL1RR029887 from the National Cancer for Research Resources (NCRR), Bethesda, MD. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIEHS or NCI, the National Institutes of Health, the Centers for Disease Control and Prevention, Atlanta, Georgia, or the California Department of Public Health.
Author Manuscript
Reference List
Author Manuscript
1. Rosner B, Colditz GA, Willett WC. Reproductive risk factors in a prospective study of breast cancer: the Nurses’ Health Study. Am J Epidemiol. 1994; 139:819–835. [PubMed: 8178795] 2. Wehkalampi K, Silventoinen K, Kaprio J, et al. Genetic and environmental influences on pubertal timing assessed by height. Am J Hum Biol. 2008; 20:417–423. [PubMed: 18293372] 3. Rogol AD, Clark PA, Roemmich JN. Growth and pubertal development in children and adolescents: effects of diet and physical activity. Am J Clin Nutr. 2000; 72:521S–528S. [PubMed: 10919954] 4. Hughes RE, Jones E. Intake of dietary fibre and the age of menarche. Ann Hum Biol. 1985; 12:325– 332. [PubMed: 2994546] 5. Koo MM, Rohan TE, Jain M, McLaughlin JR, Corey PN. A cohort study of dietary fibre intake and menarche. Public Health Nutr. 2002; 5:353–360. [PubMed: 12020388] 6. Merzenich H, Boeing H, Wahrendorf J. Dietary fat and sports activity as determinants for age at menarche. Am J Epidemiol. 1993; 138:217–224. [PubMed: 8356963] 7. Gunther AL, Karaolis-Danckert N, Kroke A, Remer T, Buyken AE. Dietary protein intake throughout childhood is associated with the timing of puberty. J Nutr. 2010; 140:565–571. [PubMed: 20042466] 8. Wiley AS. Milk intake and total dairy consumption: associations with early menarche in NHANES 1999–2004. PLoS One. 2011; 6:e14685. [PubMed: 21347271] 9. Ramezani TF, Moslehi N, Asghari G, et al. Intake of dairy products, calcium, magnesium, and phosphorus in childhood and age at menarche in the Tehran Glucose and Lipid Study. PLoS One. 2013; 8:e57696. [PubMed: 23451261] 10. Carwile JL, Willett WC, Wang M, et al. Milk Consumption after Age 9 Years Does Not Predict Age at Menarche. J Nutr. 2015; 145:1900–8. [PubMed: 26136590] 11. Moisan J, Meyer F, Gingras S. Diet and age at menarche. Cancer Causes Control. 19901:149–154. Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 8
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
12. Cheng G, Buyken AE, Shi L, et al. Nutr Rev. 2012; 70:133–152. [PubMed: 22364156] 13. Segovia-Siapco G, Pribis P, Messina M, Oda K, Sabate J. Is soy intake related to age at onset of menarche? A cross-sectional study Nutr J. 2014; 13:54.doi: 10.1186/1475-2891-13-54.:54-13 [PubMed: 24889551] 14. Mervish NA, Gardiner EW, Galvez MP, et al. Dietary flavonol intake is associated with age of puberty in a longitudinal cohort of girls. Nutr Res. 2013; 33:534–542. [PubMed: 23827127] 15. Wolff MS, Britton JA, Boguski L, et al. Environmental exposures and puberty in inner-city girls. Environ Res. 2008; 107:393–400. [PubMed: 18479682] 16. Cheng G, Remer T, Prinz-Langenohl R, et al. Relation of isoflavones and fiber intake in childhood to the timing of puberty. Am J Clin Nutr. 2010; 92:556–564. [PubMed: 20631200] 17. Adgent MA, Daniels JL, Rogan WJ, et al. Early-life soy exposure and age at menarche. Paediatr Perinat Epidemiol. 2012; 26:163–175. [PubMed: 22324503] 18. Adlercreutz H. Phytoestrogens and breast cancer. J Steroid Biochem Mol Biol. 2002; 83:113–118. [PubMed: 12650707] 19. Buck K, Zaineddin AK, Vrieling A, Linseisen J, Chang-Claude J. Meta-analyses of lignans and enterolignans in relation to breast cancer risk. Am J Clin Nutr. 2010; 92:141–153. [PubMed: 20463043] 20. Giles GG. Epidemiological investigation of prostate cancer. Methods Mol Med. 2003; 81:1–19. [PubMed: 12725111] 21. Adlercreutz H, Bannwart C, Wahala K, et al. Inhibition of human aromatase by mammalian lignans and isoflavonoid phytoestrogens in childhood and age at menarche in the Tehran Lipid and Glucose Study. J Steroid Biochem. 1993; 44:147–153. Mol Biol. 22. Setchell KD. Phytoestrogens: the biochemistry, physiology, and implications for human health of soy isoflavones. Am J Clin Nutr. 1998; 68:1333S–1346S. [PubMed: 9848496] 23. Linseisen J, Piller R, Hermann S, Chang-Claude J. Dietary phytoestrogen intake and premenopausal breast cancer risk in a German case-control study. Int J Cancer. 2004; 110:284– 290. [PubMed: 15069695] 24. Mumford SL, Sundaram R, Schisterman EF, et al. Higher urinary lignan concentrations in women but not men are positively associated with shorter time to pregnancy. J Nutr. 2014; 144:352–358. [PubMed: 24401816] 25. Biro FM, Galvez MP, Greenspan LC, et al. Pubertal Assessment Method and Baseline Characteristics in a Mixed Longitudinal Study of Girls. Pediatrics. 2010; 126(3):e583–90. [PubMed: 20696727] 26. Willett W, Stampfer MJ. Total energy intake: implications for epidemiologic analyses. Am J Epidemiol. 1986; 124:17–27. [PubMed: 3521261] 27. CDC. Overweight and Obesity. 2015. www.cdc.gov/obesity/childhood/defining.html 28. Euling SY, Selevan SG, Pescovitz OH, Skakkebaek NE. Role of environmental factors in the timing of puberty. Pediatrics. 2008; 121(Suppl 3):S167–71. S167-S171. DOI: 10.1542/peds. 2007-1813C [PubMed: 18245510] 29. Sampson L, Rimm E, Hollman PC, de Vries JH, Katan MB. Flavonol and flavone intakes in US health professionals. J Am Diet Assoc. 2002; 102:1414–1420. [PubMed: 12396158] 30. Fink BN, Steck SE, Wolff MS, Kabat GC, Gammon MD. Construction of a flavonoid database for assessing intake in a population-based sample of women on Long Island, New York. Nutr Cancer. 2006; 56:57–66. [PubMed: 17176218] 31. Biro FM, Khoury P, Morrison JA. Influence of obesity on timing of puberty. Int J Androl. 2006; 29:272–277. [PubMed: 16371114] 32. Jungbauer A, Medjakovic S. Phytoestrogens and the metabolic syndrome. J Steroid Biochem Mol Biol. 2014; 139:277–89. Epub@2013 Jan 11.:277–289. DOI: 10.1016/j.jsbmb.2012.12.009 [PubMed: 23318879] 33. Hu Y, Song Y, Franke AA, Hu FB, van Dam RM. A Prospective Investigation of the Association Between Urinary Excretion of Dietary Lignan Metabolites and Weight Change in US Women. Am J Epidemiol. 2015; 182:503–511. [PubMed: 26290574]
Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 9
Author Manuscript Author Manuscript
34. Berkey CS, Gardner JD, Frazier AL, Colditz GA. Relation of childhood diet and body size to menarche and adolescent growth in girls. Am J Epidemiol. 2000; 152:446–452. [PubMed: 10981459] 35. dos Santos Silva I, De Stavola BL, Mann V, Kuh D, Hardy R, Wadsworth ME. Prenatal factors, childhood growth trajectories and age at menarche. Int J Epidemiol. 2002; 31:405–412. [PubMed: 11980804] 36. Sloboda DM, Hart R, Doherty DA, Pennell CE, Hickey M. Age at menarche: Influences of prenatal and postnatal growth. J Clin Endocrinol Metab. 2007; 92:46–50. [PubMed: 17062767] 37. Ong KK, Ahmed ML, Emmett PM, Preece MA, Dunger DB. Association between postnatal catchup growth and obesity in childhood: prospective cohort study. BMJ. 2000; 320:967–971. [PubMed: 10753147] 38. Thanos J, Cotterchio M, Boucher BA, Kreiger N, Thompson LU. Adolescent dietary phytoestrogen intake and breast cancer risk (Canada). Cancer Causes Control. 2006; 17:1253–1261. [PubMed: 17111256] 39. Cutler GJ, Flood A, Hannan P, Neumark-Sztainer D. Major patterns of dietary intake in adolescents and their stability. J Nutr. 2009; 139:323–328. [PubMed: 19091799]
Author Manuscript Author Manuscript Pediatr Res. Author manuscript; available in PMC 2017 August 11.
Mervish et al.
Page 10
Table 1
Author Manuscript
Baseline Characteristics of 1044 BCERP participants, 2004–2014. Characteristic Mean (Std) Baseline age (years)
7.3 (.68)
Child race/ethnicity
N
%
Black
316
30
Hispanic
316
30
White
357
34
Asian
55
5
≤ High school
296
29
> High school
725
71
High school
35 29 36
High school
30 35 35
